Institutional-grade signals, published as evidence.
We operate as a quantitative research desk. Every signal is validated with Combinatorial Purged Cross-Validation and subjected to an NLP confirmation gate, with P&L reported net of IBKR and regulatory costs.
Live integrity snapshot
Methodology
Anti-hype processCPCV Validation
Thousands of alternate market paths prevent overfit and quantify robustness.
NLP Confirmatory Veto
Signals are suppressed when macro sentiment contradicts price structure.
Execution Realism
Reports include IBKR Pro fixed commissions, FINRA TAF, and modeled slippage.
Research stack
Institutional deliverablesWeekly Transparency
Drawdowns, vetoed signals, and pattern cohort pruning decisions.
Pattern Cohort Validation
CPCV path distributions and out-of-sample expectancy by asset.
API Research Feed
Parquet-backed evidence database for family offices and allocators.
Validation layer
Evidence databaseEach report is anchored to reproducible Parquet outputs. Signals, trades, CPCV paths, and NLP veto statistics are exposed through a controlled evidence layer, enabling independent verification and audit-ready review.
Awaiting CPCV runs to populate evidence metrics.
Access
Publisher modelResearch Newsletter
Weekly institutional summary and methodology updates.
Premium Signals
Daily alerts and exit sequencing delivered via email or Telegram.
Institutional API
Direct evidence feed and custom regime analytics.